But Lilley told "20/20" that the wife, who is right-handed, would have had a hard time shooting herself in her right hip area, where bullet fragments lodged.
Anderson countered that Linda used a two-handed grip.
"The fact that the bullet wound was on her right side, I believe she had the two-hand hold," Anderson said. "She used her dominant hand to steady the gun, and she used her left hand to pull the trigger -- I mean it doesn't make sense that you'd do this, you would go across your body."
Linda Dolloff denied she had shot herself.
"To me, that is the most ridiculous part of the story," she told ABC News. "I find it absurd. I don't know how many people could emotionally get themselves into a place where they could physically do that. I don't ever see myself getting into a position where I could actually point a gun at myself and pull the trigger. No."
Lilley said the evidence ruled out her having shot herself.
"We measured her arms, it's 18-and-a-half inches from the armpit to the center of her hand," he said. "And the gun is six or seven inches maybe ..."
Linda Dolloff said, "I'd be within two or three inches of my body. There would be blood in the barrel. There would be something called tattooing around the wound. There would be the burn marks on my shirt. There is no evidence of any of those, and my DNA is not on the trigger, and I have no idea [how] I could shoot myself if I did not pull the trigger."
Prosecutors knew this would be a major part of Linda Dolloff's defense. So they sent the .22 caliber Ruger handgun and the shirt Linda was wearing that night to the Maine State Crime Lab for testing.
"The question was how far was that gun from her clothing when it was fired," said forensic firearms expert Kimberly Stevens, who worked on the case. The standard procedure, Stevens said, is to look to see if there's gunpowder on the shirt.
"If the gunshot wound was self-inflicted, then we're looking at a shot that would need to be fairly close-range, because someone's arm is only so far," Stevens said. "And that, there's a distance at which the powder should be transferred."
But Linda Dolloff was right: The gunpowder the state expected to find on the shirt was not there.
"I looked at it visually, under the stereo microscope, and there was no powder residue on the shirt," Stevens said. "And even though you get a muzzle flash, there was no singeing or burning of the material on that shirt."
That evidence suggested the gun was fired from too far away for Linda to have shot herself.
But the state was not ready to give up on its theory. Stevens said gunpowder can shake loose from clothing, so it doesn't always show up in tests. She decided on a second test, this time looking not for powder but for lead vapors.
"The lead vapor is very fine-particle, it gets embedded in the weave and it's not always easy to see," Stevens said.
But if there is lead vapor on fabric, she said, a chemical spray will turn it purple and visible to the naked eye.
"So having no powder is not unusual, and having vaporous lead is very significant," Stevens said.
Stevens had tested the shirt Linda Dolloff was wearing when she was shot for lead vapors. The results were unequivocal: vapors were embedded in the fabric.
Prosecutors told the jury gunpowder wasn't important, but lead vapors indicated the gun was fired at close range.